LLMs for AI policies evaluation: discussing data sharing at supra national level
Keywords: Large Language Models, AI policy, AI governance, data sharing, United Nations
TL;DR: The paper examines the role of LLMs in analysing the impact of AI policies, highlighting the importance of data sharing and international cooperation in navigating challenges and maximizing benefits in AI governance.
Abstract: Large Language Models (LLMs) possess remarkable capabilities in both analysing vast amount of data and generating coherent human-readable output. This makes LLMs invaluable tools for various applications, and in different sectors, including policymaking.
One notable application is in sentiment analysis, where LLMs can assess the effectiveness of policies from different perspectives. By analyzing sentiment, these models can identify which policies are effective and which are not, helping policymakers make informed decisions. Additionally, LLMs can evaluate the efficacy of policies by considering trade-offs and costs, providing a comprehensive understanding of their impact.
Such an analysis of different jurisdictional experiences on specifically AI policies has great potential, given the fact that different countries are adopting different approaches. However, challenges exist. Among others, data sharing among countries is limited, hindering comprehensive analysis. To address this, an international platform such as the United Nations could facilitate data sharing and analysis.
This paper addresses the relevance of supra national data sharing in relation to the deployment of LLMs for AI policies evaluation.
Submission Number: 8
Loading